Data-dependent acquisition (DDA) (also referred to, in various commercial implementations, as Information Dependent Acquisition, Data Directed Analysis, and AUTO MS/MS) is a valuable and widely-used tool in the mass spectrometry art, particularly for the analysis of complex samples. Generally described, data-dependent acquisition involves using data derived from an experimentally-acquired mass spectrum in an “on-the-fly” manner to direct the subsequent operation of a mass spectrometer; for example, a mass spectrometer may be switched between MS and MS/MS scan modes upon detection of an ion species of potential interest. Utilization of data-dependent acquisition methods in a mass spectrometer provides the ability to make automated, real-time decisions in order to maximize the useful information content of the acquired data, thereby avoiding or reducing the need to perform multiple chromatographic runs or injections of the analyte sample. These methods can be tailored for specific desired objectives, such as enhancing the number of peptide identifications from the analysis of a complex mixture of peptides derived from a biological sample.
DDA methods may be characterized as having one or more input criteria, and one or more output actions. The input criteria employed for conventional data-dependent methods are generally based on parameters such as intensity, intensity pattern, mass window, mass difference (neutral loss), mass-to-charge (m/z) inclusion and exclusion lists, and product ion mass. The input criteria are employed to select one or more ion species that satisfy the criteria. The selected ion species are then subjected to an output action (examples of which include performing MS/MS or MSn analysis and/or high-resolution scanning). In one instance of a typical data-dependent experiment, a group of ions are mass analyzed, and ion species having mass spectral intensities exceeding a specified threshold are subsequently selected as precursor ions for MS/MS analysis, which may involve operations of isolation, dissociation of the precursor ions, and mass analysis of the product ions.
Current DDA methods are target specific and utilize the targeted precursor m/z value to direct product ion acquisition around a narrow isolation width as well as creation of dynamic exclusion (DE) values. In addition, existing DDA methods generally use static tandem mass spectral acquisition parameters (e.g., automatic gain control (AGC) target value, maximum ion fill times, resolution, etc.). This method achieves positive outcomes for MS/MS spectrum identification, but penalizes the MS/MS coverage and limits the precursor mass range coverage.
In translational/clinical research, the goal is to perform global protein profiling on a targeted proteome to identify, verify (sequencing), and quantify potential marker panels. This requires robust, accurate, and reproducible experimental liquid chromatography-mass spectrometry (LC-MS) and MS/MS data acquisition, enabling qualitative and quantitative analysis of all compounds in the biological mixture repetitively across all biological samples that are analyzed in a study. In addition, the resulting data must facilitate retrospective data mining to accommodate various hypotheses-based data processing on existing and yet-to-be determined/developed knowledge bases. These knowledge bases can be described as protein or peptide sequences, SNPs, as well as post-translational modifications.
One drawback that is associated with current DDA methods is that not all MS/MS information is available in every single injection, leaving information gaps which can make it impossible to assign a particular identification to a precursor recorded in the survey MS1 spectrum. This makes a retrospective data mining difficult, as a particular precursor of interest may be quantified from the MS1 trace, but no MS/MS information might be available for the identification/confirmation of that MS1 trace. Even with an improvement in the data analysis step using a so-called “match-between-runs,” which maps MS1 precursors across several injections to match the corresponding MS/MS information, the data set may still contain gaps that make the retrospective analysis of the data set incomplete.
It would therefore be beneficial to provide methods that overcome at least some of the above-mentioned disadvantages and/or limitations.